As is known, gas lasers have been developed in the past wherein one gas in its metastable vibrational level can be used to selectively populate an upper level of another gas through resonant transfer via inelastic collisions. One such system is described in Patel U.S. Pat. No. 3,411,105 wherein the vibrational energy of nitrogen is transferred to carbon dioxide, the active laser medium. Other systems of this type using polyatomic molecules as the active medium are disclosed in Byrne U.S. Pat. No. 3,605,038.
In the carbon dioxide-nitrogen laser, for example, an electric discharge in a mixture of these gases results in collisions of electrons with nitrogen molecules, thereby exciting them vibrationally. The cross section for these excitation processes is quite high. Since the nitrogen molecule, like all homonuclear diatomic molecules, possesses no intrinsic electric dipole moment in any of its vibrational states, relaxation of vibrationally excited nitrogen via emission of radiation is impossible. An excited nitrogen molecule, therefore, retains its excess energy until it gives it up by collision which can be either with a container wall or with some other molecular species.
It happens that the first excited vibrational state of nitrogen which lies at 2331 cm.sup.-1 above the ground vibrational state, coincides almost exactly in energy with the first excited state of the asymmetric stretching vibration of carbon dioxide at 2349 cm.sup.-1. Because of this near coincidence, excited nitrogen molecules can, upon collision with unexcited carbon dioxide molecules, efficiently transfer their energy to the carbon dioxide molecules, leaving the latter in the first excited state of the asymmetric stretch vibration. Since this excitation occurs preferentially, the first excited state of carbon dioxide becomes populated while the lower lying states remain unpopulated. As a result, a population inversion, much like that in a four-level laser, is immediately created with energy being released in the form of coherent light. Laser action is usually observed in a carbon dioxide-nitrogen laser at 10.6 microns; however, by proper cavity construction, efficient laser action can also be observed at 9.6 microns.
Although laser action will occur in carbon dioxide-nitrogen mixtures without the addition of other gases, the addition of helium or some other noble gas to the mixture increases the efficiency markedly. The helium or other noble gas acts to slow down the rate of energy loss from excited nitrogen atoms by wall collisions, moderates the energy of the discharge electrons, increases the nitrogen excitation efficiency, and spreads the discharge more uniformly throughout the active medium. Efficiencies as high as 31% have been observed in electric discharge pumped carbon dioxide-nitrogen-helium lasers, the maximum theoretical efficiency being given by the ratios of the energies of the initial laser state and of the laser quantum and is 41% for the 10.6 micron transition.
One of the best known atmospheric transmission windows lies in the near-infrared between 3 and 5 microns. Because of the low background in this region and because relatively efficient photodetectors exist for these wavelengths, many present-day infrared optical systems operate in this region. Several coherent sources are available which operate in the region between 3 and 5 microns, but most of these are low-power devices. The 3.39 micron line of the helium-neon laser, for example, gives only very low-power outputs on the order of a few milliwatts, and the alignment problems associated with this laser system prevent its use in an airborne system. Harmonic generation of carbon dioxide laser radiation can be used to generate coherent radiation at several wavelengths in the region of 3 to 5 microns; however harmonic generation at such long wavelengths is an inherently inefficient process and few harmonic generating materials have yet been discovered. Finally, parametric processes can be used to downconvert ruby or neodymium laser radiation to this region, but such processes have not as yet been made very efficient and the optical alignment problems for parametric processes are severe.